Terrace reconstruction and long profile projection: a case study from the Solent river system near Southampton, England

River terrace sequences are important frameworks for archaeological evidence and as such it is important to produce robust correlations between what are often fragmentary remnants of ancient terraces. This paper examines both conceptual and practical issues related to such correlations, using a case study from the eastern part of the former Solent River system near Southampton, England. In this region two recent terrace schemes have been constructed using different data to describe the terrace deposits: one based mainly on terrace surfaces; the other on gravel thicknesses, often not recording the terrace surface itself. The utility of each of these types of data in terrace correlation is discussed in relation to the complexity of the record, the probability of post-depositional alteration of surface sediments and comparison of straight-line projections with modern river long profiles. Correlation using age estimates is also discussed, in relation to optically stimulated luminescence dating of sand lenses within terrace gravels in this region during the PASHCC project. It is concluded that the need for replication at single sites means that this approach has limited use for correlative purposes, although dating of sediments is important for understanding wider landscape evolution and patterns of human occupation.